Various applications require detection and analog processing of optical signals. One such application includes a conventional detection circuit for optical signals that addresses the issue of common mode rejection of laser intensity noise by automatically adjusting the effective photocurrent of one of two detection diodes, and then subtracts the resulting effective photocurrents to minimize the common mode intensity noise. The photocurrent adjustment is performed by a differential pair of transistors. In the detection circuit, the photocurrent of the second photodiode is divided by 2x (where x is a variable close to one). In order for the resultant second photocurrent to equal the first photodiode current, the initial photocurrent of the second diode must be two times the photocurrent of the first photodiode. This requires that the intensity of the second beam is two times the intensity of the first beam. This constraint requires the reduction of the intensity of the first beam by optical means resulting in the reduction of detector sensitivity and a complication of the optics.
The conventional detection circuit also requires a photodiode bias voltage for each diode, where a positive bias voltage and a negative bias voltage is required. Photodiode bias voltages add complexity and noise to the photodiode generated currents. Such detection circuit only provides detection of the difference of the photo intensities and only closes the loop on an optimum direct current (DC) photo current match. Since the current servo loop is not perfectly linear, the optimum match at the signal carrier frequency is not optimized.